{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:N4EBOAOM7XUCME4IB4FVCR6EHI","short_pith_number":"pith:N4EBOAOM","canonical_record":{"source":{"id":"2412.18426","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-12-24T13:41:47Z","cross_cats_sorted":[],"title_canon_sha256":"3b826fb809e0a5dc80af08eead05ee23efe70da7c6dbb7e070ad5a64f1eb68e7","abstract_canon_sha256":"b2bfd0a0a4b3186b1514191ae3660b3b9b6401c6e2ca5c412d4f6c9340a64c54"},"schema_version":"1.0"},"canonical_sha256":"6f081701ccfde82613880f0b5147c43a2d02ddc887510b98680ff20b1947b864","source":{"kind":"arxiv","id":"2412.18426","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.18426","created_at":"2026-07-05T09:53:55Z"},{"alias_kind":"arxiv_version","alias_value":"2412.18426v1","created_at":"2026-07-05T09:53:55Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.18426","created_at":"2026-07-05T09:53:55Z"},{"alias_kind":"pith_short_12","alias_value":"N4EBOAOM7XUC","created_at":"2026-07-05T09:53:55Z"},{"alias_kind":"pith_short_16","alias_value":"N4EBOAOM7XUCME4I","created_at":"2026-07-05T09:53:55Z"},{"alias_kind":"pith_short_8","alias_value":"N4EBOAOM","created_at":"2026-07-05T09:53:55Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:N4EBOAOM7XUCME4IB4FVCR6EHI","target":"record","payload":{"canonical_record":{"source":{"id":"2412.18426","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-12-24T13:41:47Z","cross_cats_sorted":[],"title_canon_sha256":"3b826fb809e0a5dc80af08eead05ee23efe70da7c6dbb7e070ad5a64f1eb68e7","abstract_canon_sha256":"b2bfd0a0a4b3186b1514191ae3660b3b9b6401c6e2ca5c412d4f6c9340a64c54"},"schema_version":"1.0"},"canonical_sha256":"6f081701ccfde82613880f0b5147c43a2d02ddc887510b98680ff20b1947b864","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:53:55.328598Z","signature_b64":"g81zlx1KZVaVUsT7OROrC8pPlQcANa/UwCXgwRlnuMch/RcCDyYpgdzVGPE9KCFw2oNVHkUbUIa2yVA1ChkEBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6f081701ccfde82613880f0b5147c43a2d02ddc887510b98680ff20b1947b864","last_reissued_at":"2026-07-05T09:53:55.328099Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:53:55.328099Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.18426","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:53:55Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"lbCB0Ik0/CYHLQUlIxKtHqjxypP7vBiRIoqWBOVPvL0rmLlKfxxeejX7MX0MMUYkVBwJBhU2whbqj+Bp41lJAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-21T18:51:16.764983Z"},"content_sha256":"999f7a2416592e43445365216f5a9394f84ab98d4903359701a8ee4a40932eb0","schema_version":"1.0","event_id":"sha256:999f7a2416592e43445365216f5a9394f84ab98d4903359701a8ee4a40932eb0"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:N4EBOAOM7XUCME4IB4FVCR6EHI","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"GUI Testing Arena: A Unified Benchmark for Advancing Autonomous GUI Testing Agent","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Haozhan Shen, Jiahui Song, Jianwei Yin, Kangjia Zhao, Leigang Sha, Tiancheng Zhao, Xiubo Liang, Zhi Chen","submitted_at":"2024-12-24T13:41:47Z","abstract_excerpt":"Nowadays, research on GUI agents is a hot topic in the AI community. However, current research focuses on GUI task automation, limiting the scope of applications in various GUI scenarios. In this paper, we propose a formalized and comprehensive environment to evaluate the entire process of automated GUI Testing (GTArena), offering a fair, standardized environment for consistent operation of diverse multimodal large language models. We divide the testing process into three key subtasks: test intention generation, test task execution, and GUI defect detection, and construct a benchmark dataset b"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.18426","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2412.18426/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:53:55Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"edaRSJOGVgpEPgKNFkp25EdnPfxLpY05aNLy8Ka3bDyIRLcrRR7gZELqqFVwUA3Rc32cmU424xHGBjBz2717DA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-21T18:51:16.765764Z"},"content_sha256":"80daffbab066df1aa2b8cd20d4106e2d7c59403cd6c62b67df291b65353a4675","schema_version":"1.0","event_id":"sha256:80daffbab066df1aa2b8cd20d4106e2d7c59403cd6c62b67df291b65353a4675"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/N4EBOAOM7XUCME4IB4FVCR6EHI/bundle.json","state_url":"https://pith.science/pith/N4EBOAOM7XUCME4IB4FVCR6EHI/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/N4EBOAOM7XUCME4IB4FVCR6EHI/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-21T18:51:16Z","links":{"resolver":"https://pith.science/pith/N4EBOAOM7XUCME4IB4FVCR6EHI","bundle":"https://pith.science/pith/N4EBOAOM7XUCME4IB4FVCR6EHI/bundle.json","state":"https://pith.science/pith/N4EBOAOM7XUCME4IB4FVCR6EHI/state.json","well_known_bundle":"https://pith.science/.well-known/pith/N4EBOAOM7XUCME4IB4FVCR6EHI/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:N4EBOAOM7XUCME4IB4FVCR6EHI","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"b2bfd0a0a4b3186b1514191ae3660b3b9b6401c6e2ca5c412d4f6c9340a64c54","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-12-24T13:41:47Z","title_canon_sha256":"3b826fb809e0a5dc80af08eead05ee23efe70da7c6dbb7e070ad5a64f1eb68e7"},"schema_version":"1.0","source":{"id":"2412.18426","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.18426","created_at":"2026-07-05T09:53:55Z"},{"alias_kind":"arxiv_version","alias_value":"2412.18426v1","created_at":"2026-07-05T09:53:55Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.18426","created_at":"2026-07-05T09:53:55Z"},{"alias_kind":"pith_short_12","alias_value":"N4EBOAOM7XUC","created_at":"2026-07-05T09:53:55Z"},{"alias_kind":"pith_short_16","alias_value":"N4EBOAOM7XUCME4I","created_at":"2026-07-05T09:53:55Z"},{"alias_kind":"pith_short_8","alias_value":"N4EBOAOM","created_at":"2026-07-05T09:53:55Z"}],"graph_snapshots":[{"event_id":"sha256:80daffbab066df1aa2b8cd20d4106e2d7c59403cd6c62b67df291b65353a4675","target":"graph","created_at":"2026-07-05T09:53:55Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2412.18426/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Nowadays, research on GUI agents is a hot topic in the AI community. However, current research focuses on GUI task automation, limiting the scope of applications in various GUI scenarios. In this paper, we propose a formalized and comprehensive environment to evaluate the entire process of automated GUI Testing (GTArena), offering a fair, standardized environment for consistent operation of diverse multimodal large language models. We divide the testing process into three key subtasks: test intention generation, test task execution, and GUI defect detection, and construct a benchmark dataset b","authors_text":"Haozhan Shen, Jiahui Song, Jianwei Yin, Kangjia Zhao, Leigang Sha, Tiancheng Zhao, Xiubo Liang, Zhi Chen","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-12-24T13:41:47Z","title":"GUI Testing Arena: A Unified Benchmark for Advancing Autonomous GUI Testing Agent"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.18426","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:999f7a2416592e43445365216f5a9394f84ab98d4903359701a8ee4a40932eb0","target":"record","created_at":"2026-07-05T09:53:55Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"b2bfd0a0a4b3186b1514191ae3660b3b9b6401c6e2ca5c412d4f6c9340a64c54","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-12-24T13:41:47Z","title_canon_sha256":"3b826fb809e0a5dc80af08eead05ee23efe70da7c6dbb7e070ad5a64f1eb68e7"},"schema_version":"1.0","source":{"id":"2412.18426","kind":"arxiv","version":1}},"canonical_sha256":"6f081701ccfde82613880f0b5147c43a2d02ddc887510b98680ff20b1947b864","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6f081701ccfde82613880f0b5147c43a2d02ddc887510b98680ff20b1947b864","first_computed_at":"2026-07-05T09:53:55.328099Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:53:55.328099Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"g81zlx1KZVaVUsT7OROrC8pPlQcANa/UwCXgwRlnuMch/RcCDyYpgdzVGPE9KCFw2oNVHkUbUIa2yVA1ChkEBw==","signature_status":"signed_v1","signed_at":"2026-07-05T09:53:55.328598Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.18426","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:999f7a2416592e43445365216f5a9394f84ab98d4903359701a8ee4a40932eb0","sha256:80daffbab066df1aa2b8cd20d4106e2d7c59403cd6c62b67df291b65353a4675"],"state_sha256":"09ba0c9ae532c2f245206e7dbaa994cf6290f5cb04bacd04925d60530673bacb"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KwHz6tah8dLSXMb2RjpJ35CX5XVXjaFL/fWMs/3aPyg9DSHw3ITavk+EPHU4J6QMg416WwpRQievpvLcf0NqAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-21T18:51:16.771781Z","bundle_sha256":"32b1123123a29fe812ec357b19994b9dc23bdc9d45de4c3a85a3aa237bb31835"}}